Contemporary research into facial recognition technology began in earnest in the mid-1990s at the behest of DARPA, the Defense Advanced Research Projects Agency. The military wanted facial recognition to exist, so DARPA began funding researchers in computer science and computer vision to work on the problem.
The military realized that to do facial recognition, researchers would need to have access to thousands of images of peoples’ faces. So in the early 1990s the military funded the creation of something called the FERET database, which included tens of thousands of pictures of many thousands of people, most of whom worked at a military base in Maryland. This work was made by combing through the FERET database over many months to “curate” a selection of portraits, to retouch and color correct them, and run them through an algorithm that identifies the keypoints in their faces. One of the ways I think about these portraits is as a kind of super-structuralism in the sense that they are images that were not made for human eyes. They are made for machine eyes. What’s more, these photos represent some of the “original-faces” of facial recognition – the “Adam and Eves” that nearly all subsequent facial recognition research has been built upon.